JP2007081755A - Image data conversion apparatus - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide an image data conversion apparatus capable of enhancing the improvement effect of image quality for smoothing a motion of a moving picture. <P>SOLUTION: When a sequence detection means 1 detects at which position of one sequence a telecine signal, being an interlace image data resulting from converting original progressive image data by the 2-3 pulldown system, exists, a memory control means 44 writes the interlace image data to frame memories 41 to 43 on the basis of a result of the detection and reads inverse conversion image data from the frame memories, an interpolation means 45 composes the two inverse conversion image data consecutively read at a rate of causing a change in the respective components by 2/5 each to produce four interpolation frame image data, and a selection means 46 selectively extracts the one inverse conversion image data read from the frame memories and the four interpolation frame image data produced by the interpolation means in a prescribed order to form progressive image data. <P>COPYRIGHT: (C)2007,JPO&INPIT

Description

  The present invention relates to an image data conversion apparatus for converting interlaced image data obtained by converting original progressive image data of a film source by a 2-3 pulldown method into progressive image data having a frame frequency equal to that of the interlaced image data.

  When a movie image recorded on a film is recorded on a recording medium such as a DVD (Digital Versatile Disc) or transmitted by digital broadcasting or the like, a movie image of 24 frames per second has a frame frequency of 24 Hz and one frame. The number of lines (scanning lines) is recorded or transmitted on a recording medium as, for example, 525 progressive image data.

  In a playback device such as a DVD player or a receiving device such as a digital broadcast receiver, an original progressive image data having a frame frequency of 24 Hz is converted to a field frequency of 60 Hz by a 2-3 pull-down method in a telecine conversion device or an MPEG decoder. Convert to interlaced image data. This image data conversion method will be described below with reference to FIG.

  As shown in FIG. 6 (a), original progressive image data (hereinafter, image data for one frame or image data for one field is also simply referred to as an image) are frame images A, B, and C having a frame frequency of 24 Hz. , D, E, and F. When interlaced image data having a frequency of 60 Hz is generated from the original progressive image data, as shown in FIG. 6B, the odd lines of the frame image A of the original progressive image data are used as odd field images At. Are sequentially extracted as the even field image Ab, and the odd line of the frame image B of the original progressive image data is set as the odd field image Bt, the even line as the even field image Bb, and the odd line as the odd field image Bt again. Sequentially extracted, the even line of the frame image C of the original progressive image data is sequentially extracted as the even field image Cb, the odd line as the odd field image Ct, and the even line of the frame image D of the original progressive image data is even field. As the image Db, odd lines are extracted as odd field images Dt, and even lines are sequentially extracted as even field images Db. Similarly, the odd lines of the frame image E of the original progressive image data are extracted as odd fields. As the image Et, the even lines are sequentially extracted as the even field images Eb, the odd lines of the frame image F of the original progressive image data are set as the odd field images Ft, the even lines are set as the even field images Fb, and again the odd lines. Are sequentially extracted as odd field images Ft.

  Next, the odd-numbered field image At and the even-numbered field image Ab of the interlaced image data are read twice, and are simply combined to generate the frame images A and A of the inverse-transformed image data. For the field image Bt, the even field image Bb, and the odd field image Bt, adjacent field images are sequentially read and simply combined to generate frame images B, B, and B of inversely transformed image data. Similarly, frame images C, C, D, D, D, E, E, F, F, and F are generated by synthesizing inverse parity images of the inversely converted image data. As a result, as shown in FIG. 6C, inversely converted image data of 60 Hz is obtained.

  However, in the image data conversion method described above, as shown in FIG. 6C, after two frame images A are output, three frame images B are output, and thereafter the same output state is repeated. As described above, since the case where the same frame continues for two frames and the case where another same frame continues for three frames appear alternately, the motion image may not be smooth. Therefore, a plurality of image data conversion methods that smooth the motion of moving images have been proposed (see, for example, Patent Document 1 below).

  FIG. 7 shows a plurality of methods for obtaining progressive image data having a frame frequency of 60 Hz by further interpolating the inversely transformed image data shown in FIG. 6C so that the motion of the moving image looks smooth. It is a schematic diagram. Here, if the first method among the first to fifth methods shown in FIG. 7D is described, the second to fifth methods can be easily understood from the description in the drawing. Therefore, only the first method will be described, and descriptions of the second to fifth methods will be omitted.

In the first method, as shown as progressive image data (d-1), in the first frame of the progressive image data, the image A of the first frame of the inversely converted image data is used as the image of the frame as it is, and the progressive image data is displayed. In the second frame and the third frame of data, an image obtained by combining the image A of the second frame and the image B of the third frame of the inversely converted image data at a ratio of (1/2) :( 1/2), respectively. In the fourth frame of the progressive image data, the fourth frame image B of the inversely converted image data is directly used as the image of the frame, and in the fifth frame of the progressive image data, the number of the inversely converted image data An image obtained by synthesizing the image B of 5 frames and the image C of the sixth frame at a ratio of (1/2) :( 1/2) And an image, the following five frames as one sequence (service cycle) to repeat the same interpolation processing.
JP 2004-15700 A

  However, in the first to fifth image data conversion methods described in Patent Document 1 described above, when converting the inversely converted image data shown in FIG. 7C to the progressive image data shown in FIG. The composite ratio of the inversely converted image data is fixed to 1/2, changed to 1/2, 1/4, 3/4 depending on the frame position of the progressive image data, or 1/2, 1/3, 2 / 3. As shown in FIG. 7C, for example, the frame image A is output after 2 frames, and then the frame image B is output 3 frames, and thereafter the inverse output image data is repeatedly output in the same manner. It is considered that the synthesis ratio was determined appropriately. As a result, the amount of change in the composition ratio between adjacent frames varies from frame to frame, and there is a problem that the effect of improving the image quality that smoothes the motion of the moving image is not sufficient.

  The present invention has been made to solve the above-described problems, and an object of the present invention is to provide interlaced image data obtained by converting original progressive image data of a film source by a 2-3 pulldown method. An object of the present invention is to provide an image data conversion device capable of further enhancing the image quality improvement effect of smoothing the motion of a moving image when converting to progressive image data having the same frequency.

In order to achieve the above object, the present invention provides interlaced image data of 60 fields / second in which film image data of 24 frames / second obtained by converting a film source into an electrical signal is converted by a 2-3 pulldown method. In an image data converter for converting to progressive image data of 60 frames / second,
The interlaced image data for at least three frames of the film image data is written, and the odd field data and the even field data in the interlaced image data are synthesized for each frame of the film image data to have the same image quality as the film image data. A frame memory controlled to output the inverse transformed image data of
In the interlaced image data, two consecutive fields in which the first frame of the film image data is converted and three consecutive fields in which the second frame of the film image data is converted are combined. Sequence detection means for detecting which of the five fields present in the one sequence the interlaced image data is sequentially input with five fields as one sequence;
A control signal for sequentially writing the interlaced image data for at least three consecutive frames of the film image data in the frame memory and selectively reading the written interlaced image data based on the detection result of the sequence detecting means Memory control means for sequentially reading out the inversely transformed image data from the frame memory by:
Based on the detection result of the sequence detection means, the two inversely transformed image data read out continuously from the frame memory are combined and interpolated at a rate at which each image data component changes by 2/5. Interpolation means for generating frame image data;
Based on the detection result of the sequence detection means, one of the inversely transformed image data read from the frame memory is included in a position corresponding to any one field in the one sequence, and is generated by the interpolation means The four interpolated frame image data obtained are included in positions corresponding to the other four fields in the one sequence, respectively, and the time series of the inversely transformed image data and the interpolated frame image data included in the one sequence A selection means for selectively taking out from the frame memory and the interpolation means and outputting as the progressive data so that the components of the inversely transformed image data change by 2/5 each;
It is characterized by having.

The present invention also provides a 60-frame / second progressive image obtained by converting 60-field / second interlaced image data obtained by converting film image data of 24 frames / second obtained by converting a film source into an electrical signal using a 2-3 pull-down method. In an image data converter for converting data,
The interlaced image data for at least three frames of the film image data is written, and the odd field data and the even field data in the interlaced image data are synthesized for each frame of the film image data to have the same image quality as the film image data. A frame memory controlled to output the inverse transformed image data of
In the interlaced image data, two consecutive fields in which the first frame of the film image data is converted and three consecutive fields in which the second frame of the film image data is converted are combined. Sequence detection means for detecting which of the five fields present in the one sequence the interlaced image data is sequentially input with five fields as one sequence;
A control signal for sequentially writing the interlaced image data for at least three consecutive frames of the film image data in the frame memory and selectively reading the written interlaced image data based on the detection result of the sequence detecting means Memory control means for sequentially reading out the inversely transformed image data from the frame memory by:
A motion for detecting a motion vector between the two inversely transformed image data for each of the two inversely transformed image data continuously read from the frame memory based on the detection result of the sequence detecting means. Detection means;
Based on the detection result of the sequence detection means, the mutual motion vectors detected by the motion detection means are internally divided at a rate at which each component changes by 2/5, and generated by this internal division. Motion compensation means for motion-compensating and interpolating the inversely-transformed image data with two motion vector components as motion vectors for each of them, and averaging the obtained two motion compensation interpolation data to generate interpolated frame image data;
Based on the detection result of the sequence detection means, the one inversely transformed image data read from the frame memory is included in a position corresponding to any one of the one sequence, and is generated by the motion compensation means. The four interpolated frame image data are included in positions corresponding to the other four of the one sequence, respectively, and the time-series order of the inversely transformed image data and the interpolated frame image data included in the one sequence Selecting means for selectively taking out from the frame memory and the motion compensation means and outputting as the progressive data so that the components of the inversely transformed image data sequentially change by 2/5,
It is characterized by having.

Here, the motion detection means divides the inverse transformed image data into a plurality of blocks, detects a motion vector for each of the blocks,
The motion compensation means generates two motion compensation block data from the two motion vectors generated by dividing the motion vector at a rate at which each component changes by 2/5, and the motion compensation block data Are interpolated block data, and the interpolated block data synthesized for the entire frame is the interpolated frame image data.
It is characterized by that.

  According to the present invention configured as described above, reverse-converted image data equivalent to the original progressive image data of the film source is generated from the interlaced image data converted by the 2-3 pull-down method, and the reverse-converted image data Since the interpolated frame image data is generated using, it is possible to sequentially change the inversely converted image data component included in each frame of the progressive image data by 2/5, thereby smoothing the motion of the moving image. The effect of improving the image quality can be further enhanced.

  Hereinafter, the present invention will be described in detail based on preferred embodiments shown in the drawings. FIG. 1 is a schematic diagram showing a first image data conversion method implemented by the image data conversion apparatus according to the present invention. FIG. 1A shows original progressive image data of a film source, and the original progressive image data is converted by the 2-3 pull down method to form interlaced image data shown in FIG. The image data conversion apparatus according to the present invention receives the interlaced image data shown in FIG. 1 (b) as an input, first generates the inversely converted image data shown in FIG. 1 (c), and then uses the inversely converted image data. The progressive image data shown in FIG. 1 (d) is generated. For the sake of easy understanding, the configuration of each image data shown in FIGS. The embodiment will be described.

  As shown in FIG. 1A, the original progressive image data is composed of frame images A, B, C, D, E, and F having a frame frequency of 24 Hz. When interlaced image data having a frequency of 60 Hz is generated from the original progressive image data, as shown in FIG. 1B, the odd lines of the frame image A of the original progressive image data are used as odd field images At. Are sequentially extracted as the even field image Ab, and the odd line of the frame image B of the original progressive image data is set as the odd field image Bt, the even line as the even field image Bb, and the odd line as the odd field image Bt again. Sequentially extracted, the even line of the frame image C of the original progressive image data is sequentially extracted as the even field image Cb, the odd line as the odd field image Ct, and the even line of the frame image D of the original progressive image data is even field. As the image Db, odd lines are extracted as odd field images Dt, and even lines are sequentially extracted as even field images Db. Similarly, the odd lines of the frame image E of the original progressive image data are extracted as odd fields. As the image Et, the even lines are sequentially extracted as the even field images Eb, the odd lines of the frame image F of the original progressive image data are set as the odd field images Ft, the even lines are set as the even field images Fb, and again the odd lines. Are sequentially extracted as odd field images Ft.

  Next, the odd-numbered field image At and the even-numbered field image Ab of the interlaced image data shown in FIG. 1B are synthesized to generate the first frame image A of the inversely transformed image data. The field image Bt and the even field image Bb are combined to generate the image B of the second frame of the inversely converted image data. Thereafter, similarly, the inverse parity image of the inversely transformed image data is synthesized to generate the images C, D, E, and F of the third to sixth frames. As a result, as shown in FIG. 1C, 24 Hz inversely converted image data equivalent to the original progressive image data is obtained.

  Next, as shown in FIG. 1D, in the first frame of the progressive image data, the first frame image A of the inversely converted image data is used as the image of the frame as it is, and in the second frame of the progressive image data. The image obtained by combining the image A of the first frame and the image B of the second frame of the inversely converted image data at a ratio of (3/5) :( 2/5) is used as the image of the frame, and the progressive image data In 3 frames, an image obtained by combining the image A of the first frame and the image B of the second frame of the inversely converted image data at a ratio of (1/5) :( 4/5) is used as an image of the frame, and a progressive image is obtained. In the fourth frame of data, an image obtained by combining the image B of the second frame and the image C of the third frame of the inversely converted image data at a ratio of (4/5) :( 1/5) is the image of the frame. In the fifth frame of the progressive image data, an image obtained by combining the second frame image B and the third frame image C of the inversely converted image data at a ratio of (2/5) :( 3/5). The same interpolation process is repeated with the following 5 frames as one sequence. These interpolation processes are executed in synchronism with the field images At, Ab, Bt, Bb, Bt, Cb,..., Ft of the interlaced image data shown in FIG. Progressive image data is obtained.

  FIG. 2 is a schematic diagram showing a second image data conversion method implemented by the image data conversion apparatus according to the present invention. Here, the conversion process between FIGS. 2 (a), (b) and (c) is exactly the same as the conversion process between FIGS. 1 (a), (b) and (c), and the description thereof will be omitted. Only the case where the inversely converted image data shown in FIG. 2C is converted into the progressive image data shown in FIG. 2D will be described.

  As shown in FIG. 2D, in the first frame of the progressive image data, the image A of the first frame and the image B of the second frame of the inversely converted image data are (4/5): (1/5). In the second frame of the progressive image data, the image A of the first frame and the image B of the second frame of the inversely converted image data are (2/5) :( 3 / 5) is used as the image of the corresponding frame, and in the third frame of the progressive image data, the second frame image B of the inversely converted image data is used as it is as the image of the corresponding frame, and the progressive image data In 4 frames, an image obtained by combining the image B of the second frame and the image C of the third frame of the inversely converted image data at a ratio of (3/5) :( 2/5) is used as the image of the frame, and In the fifth frame of the progressive image data, an image obtained by combining the image B of the second frame and the image C of the third frame of the inversely converted image data at a ratio of (1/5) :( 4/5) The same interpolation process is repeated with an image and 5 frames as one sequence. These interpolation processes are executed in synchronization with the field images At, Ab, Bt, Bb, Bt, Cb,..., Ft of the interlaced image data shown in FIG. Progressive image data is obtained.

  As described above, the progressive image data converted by the first image data conversion method shown in FIG. 1 and the second image data conversion method shown in FIG. 2 are inversely converted image data A, B, C, D, E, respectively. , F components are arranged so as to sequentially change by 2/5 for each frame. By performing such interpolation processing, the effect of improving the image quality that smoothes the motion of the moving image is enhanced.

Next, an embodiment of the image data conversion apparatus according to the present invention will be described.
<First Embodiment>
FIG. 3 is a block diagram showing the configuration of the first embodiment of the image data conversion apparatus according to the present invention. In the first embodiment, when an input interlace signal is a non-telecine signal, an ordinary “interlace → progressive conversion process” is performed to obtain an output progressive signal. On the contrary, when the input interlace signal is the interlace image data shown in FIG. 1B called a telecine signal, the progressive image data shown in FIG. 1D is output as an output progressive signal.

  In FIG. 3, an input interlace signal is applied to a sequence detection circuit 1 and a switch (abbreviated as sw in the drawing) 2. Among these, the sequence detection circuit 1 detects that the input interlace signal is a telecine signal converted by the 2-3 pull-down method. That is, the sequence detection circuit 1 includes two consecutive fields in which the original progressive image data of the first frame of the film source is converted, and three consecutive fields in which the original progressive image data of the second frame of the film source is converted. Are detected, and the interlaced image data that is sequentially input is detected at which position in one sequence. The switch 2 outputs signals from different terminals depending on whether or not the interlaced image data is a telecine signal according to the detection result of the sequence detection circuit 1. The input terminal of the normal progressive conversion circuit 3 is connected to one output terminal of the switch 2, that is, the output terminal of the non-telecine signal, and the telecine signal progressive conversion circuit 4 is connected to the other output terminal, that is, the output terminal of the telecine signal. The input end is connected. Normally, the progressive conversion circuit 3 adjusts the output timing of progressive image data input thereto, and the telecine signal progressive conversion circuit 4 is based on the detection result of the sequence detection circuit 1 to convert the inversely converted image of 24 Hz from the interlaced image data. Data is generated, and progressive image data of 60 Hz is generated from the inversely converted image data. The output terminal of the normal progressive conversion circuit 3 and the output terminal of the telecine signal progressive conversion circuit 4 are connected to the input terminal of the selection circuit 6, respectively. Based on the detection result of the sequence detection circuit 1, the selection circuit 6 selects the output of the normal progressive conversion circuit 3 when the input interlace signal is not a telecine signal, and the output of the telecine signal progressive conversion circuit 4 when the input interlace signal is a telecine signal. Are selected as output progressive signals.

  The telecine signal progressive conversion circuit 4 described above writes interlaced image data for three frames of the film source, synthesizes odd field data and even field data for each frame of the film source, and reverses the same as the original progressive image data. Based on the detection results of the frame memories 41, 42, 43 for generating the converted image data and the sequence detection circuit 1, the interlaced image data for three consecutive frames of the film source are sequentially stored in the frame memories 41, 42, 43. Based on the detection result of the sequence detection circuit 1 and the memory control unit 44 that sequentially reads the inversely converted image data generated in the frame memories 41, 42, and 43, and reads the frame memory 41, 42, and 43 sequentially. The two inverse transformed image data Based on the detection result of the interpolating circuit 45 that generates interpolated frame image data by synthesizing at a rate that these components change by 2/5, and read from the frame memories 41, 42, and 43 based on the detection result of the sequence detecting circuit 1 One inversely transformed image data is included in a position corresponding to any one of one sequence, and the four interpolated frame image data generated by the interpolation circuit 45 are in positions corresponding to the other four in one sequence, respectively. Select from the frame memories 41, 42, and 43 and the interpolation circuit 45 that the order of the inversely converted image data and the interpolated frame image data included in one sequence changes the inversely converted image data components by 2/5 each. And a selection circuit 46 for taking out and outputting as progressive data.

  The operation of the first embodiment configured as described above will be described below. First, the input interlace signal is input to the sequence detection circuit 1 to detect whether it is a telecine signal converted by the 2-3 pull-down method. Whether the position is interlaced image data is detected. The switch 2 applies the input interlace signal to the normal progressive conversion circuit 3 when it is detected that the input interlace signal is not a telecine signal, and converts the interlace image data to the telecine signal progressive conversion when it is detected that the input interlace signal is a telecine signal. Add to circuit 4. When the input interlace signal is determined not to be a telecine signal in the sequence detection circuit 1, the normal progressive conversion circuit 3 performs “interlace → progressive conversion processing” on the input signal and outputs it. In this process, for example, interpolation is performed on the screen from the input field image to generate a parity field opposite to the current field, and a progressive signal is generated by composing with the original field to form a frame. . The telecine signal progressive conversion circuit 4 generates inversely converted image data from the interlaced image data, and further generates and outputs progressive image data from the inversely converted image data. Based on the detection result of the sequence detection circuit 1, the selection circuit 6 selects the output of the normal progressive conversion circuit 3 when the input interlace signal is not a telecine signal, and the output of the telecine signal progressive conversion circuit 4 when the input interlace signal is a telecine signal. To select an output progressive signal.

  Next, the detailed operation of the telecine signal progressive conversion circuit 4 will be described. When the sequence detection circuit 1 determines that the input interlace signal is a telecine signal, the field image data At and Ab shown in FIG. The field image data Bt and Bb shown in FIG. 1B are written into the frame memory 42, and the field image data Cb and Ct shown in FIG. 1B are written into the frame memory 43. The frame memories 41, 42, and 43 have a function of combining the two field image data written in each of them, and the frame image data A, B, and C shown in FIG. The data is read out two by two and added to the interpolation circuit 45 and the selection circuit 46.

  As shown in FIG. 1D, the interpolation circuit 45 synthesizes the image A of the first frame and the image B of the second frame of the inversely converted image data at a ratio of (3/5) :( 2/5). Next, the image A of the first frame and the image B of the second frame of the inversely transformed image data are combined at a ratio of (1/5) :( 4/5), and then the inversely transformed image data The second frame image B and the third frame image C are combined at a ratio of (4/5) :( 1/5), and then the second frame image B and the third frame of the inversely converted image data are combined. Are combined at a ratio of (2/5) :( 3/5). As shown in FIG. 1D, the selection circuit 46 arranges the image data A called from the frame memory 41 in the first frame of one sequence and sequentially combines the four images obtained by the interpolation circuit 45. The data is selected so as to be arranged in the second to fifth frames and added to the selection circuit 6.

  Next, the field image data Cb and Ct shown in FIG. 1B are transferred to the frame memory 41 by the write command of the memory control unit 44, and the field image data Db and Dt shown in FIG. The field image data Et and Eb shown in FIG. 1B are written into the frame memory 43. In the same manner as described above, frame image data C, D, and E are read from the frame memories 41, 42, and 43, and interpolation processing and selection of the next one sequence of frame image data are executed.

  Here, the operation principle of the sequence detection circuit 1 will be described. When the absolute value of the difference between the luminance signals of the current field and the field two fields before is integrated over the entire screen with respect to the interlaced image data as shown in FIG. 1B, and the value is below a predetermined threshold value When it is determined that the current field is a repetition field, and it is detected that the repetition field appears in a five-field period, it is determined that the signal is a telecine signal, and one sequence of five consecutive fields from the next field of the repetition field. And

  The case where the first image data conversion method shown in FIG. 1 is executed has been described above. However, in the second image data conversion method shown in FIG. If four image data are generated by performing interpolation processing and the selection circuit 46 calls the image data from the frame memory 41 and the selection order of the four image data synthesized by the interpolation circuit 45 is changed, it is easily implemented. The explanation is omitted.

  Thus, according to the first embodiment, it is possible to implement the first image data conversion method shown in FIG. 1 and the second image data conversion method shown in FIG. The effect of improving the image quality that smoothes the motion of the moving image can be enhanced over that of the conventional apparatus.

<Second Embodiment>
FIG. 4 is a block diagram showing the configuration of the second embodiment of the image data conversion apparatus according to the present invention. In the figure, the same reference numerals as those in the first embodiment denote the same elements. In the first embodiment, the telecine signal progressive conversion circuit 4 including the frame memories 41, 42, 43, the memory control unit 44, the interpolation circuit 45, and the selection circuit 46 is provided. However, in the second embodiment, Instead, a telecine signal progressive conversion circuit 5 including a frame memory 51, 52, 53, a memory control unit 54, a motion detection circuit 55, a motion compensation circuit 56, and a selection circuit 57 is provided. Therefore, only the telecine signal progressive conversion circuit 5 having a configuration different from that of the first embodiment will be described in the case of performing the first image data conversion method shown in FIG.

  When the sequence detection circuit 1 determines that the input interlace signal is a telecine signal, the interlace image data At and Ab shown in FIG. Interlaced image data Bt and Bb shown in FIG. 1B are written into the frame memory 52, and field image data Cb and Ct shown in FIG. 1B are written into the frame memory 53. The frame memories 51, 52, and 53 have a function of combining two interlaced image data written to each by a read control signal, and the inversely converted image data shown in FIG. 1C by the read control signal of the memory control unit 54. A, B, and C are read out two by two and added to the motion detection circuit 55, motion compensation circuit 56, and selection circuit 57.

The motion detection circuit 55 detects a motion vector for each detection area using the current frame image data, for example, B as a reference image and the past frame image data, for example, A as a reference image.
The detected motion vectors are internally divided at a predetermined ratio based on the detection result of the sequence detection circuit 1, and two internal motion vectors are calculated. The calculation result is transmitted to the motion compensation circuit 56. The motion compensation circuit 56 performs motion compensation interpolation from each frame, using one of these internally divided motion vectors as a motion vector from a past frame and the other as a motion vector from the current frame. The average is used as the compensated image output. Then, in the selection circuit 57, for example, according to the detection result of the sequence detection circuit 1, for example, the inverse transformed image data output from the frame memory 51, and four compensated image outputs subjected to the compensation interpolation by the motion compensation circuit 56, Is selected and output as progressive image data.

Specifically, when the inversely converted image data A of FIG. 1C is synthesized in the frame memory 51 and the inversely transformed image data B of FIG. 1C is synthesized in the frame memory 52, FIG. As shown in (d), in the first frame of the progressive image data, the inversely transformed image data A becomes the image of the frame as it is, but in the second frame of the progressive image data, the interpolated image shown in FIG. Become. This interpolated image is generated as described below. That is, when the motion vector from the image A to the image B is MV, the two internal motion vectors MVA and MVB obtained by internally dividing MV by 2: 3 are as follows.
MVA = (2/5) × MV
MVB = − (3/5) × MV
Then, the motion compensation interpolation data obtained by motion compensation of the image A with the internal motion vector MVA and the motion compensation interpolation data obtained by motion compensation of the image B with the internal motion vector MVB are obtained. The interpolated image data is used as the second frame image.

Similarly, for the third frame, the motion vector MV from the image A to the image B is internally divided by 4: 1, and the following two internal motion vectors MVA and MVB are obtained.
MVA = (4/5) × MV
MVB =-(1/5) * MV
Then, the motion compensation interpolation data obtained by motion compensation of the image A with the internal motion vector MVA and the motion compensation interpolation data obtained by motion compensation of the image B with the internal motion vector MVB are obtained. The interpolated image data is used as the third frame image.

  For the fourth and fifth frames, motion vectors are obtained for the inversely transformed image data B synthesized in the frame memory 52 and the inversely transformed image data C synthesized in the frame memory 53, and the motion vectors are set to 1: Progressive image data can be obtained by performing motion compensation interpolation using motion vectors internally divided by 2: 2 and 3: 2.

  The case where the first image data conversion method shown in FIG. 1 is performed has been described above. However, the second image data conversion method shown in FIG. Four image data are generated by changing the internal ratio of the motion vector and performing motion compensation interpolation by the motion compensation circuit 56, and the selection circuit 57 calls the inversely transformed image data from the frame memory 52 and compensates by the motion compensation circuit 56. Since it can be easily carried out by changing the selection order of the four interpolated image data, description thereof will be omitted.

  In this way, the second embodiment can also implement the first image data conversion method shown in FIG. 1 and the second image data conversion method shown in FIG. The effect of improving the image quality that smoothes the movement of the image can be enhanced over that of the conventional apparatus.

<Third Embodiment>
In the second embodiment, the motion vector is detected for each inversely transformed image data, but the motion detection circuit 55 divides the inversely transformed image data into a plurality of blocks, detects the motion vector for each of the blocks, The motion compensation circuit 56 divides the motion vector obtained by the motion detection circuit 55 into two motion compensation block data from the two internal motion vectors generated by dividing the motion vector at a rate at which each component changes by 2/5. It is also possible to generate and average the motion compensation block data as interpolation block data, and to combine the interpolation block data for the entire frame into interpolation frame image data. Even with this modification, the image quality improvement effect of smoothing the motion of the moving image can be enhanced over that of the conventional apparatus, as in the first and second embodiments.

It is a schematic diagram which shows the 1st image data conversion method implemented with the image data converter which concerns on this invention. It is a schematic diagram which shows the 2nd image data conversion method implemented with the image data converter which concerns on this invention. 1 is a block diagram showing a configuration of a first embodiment of an image data conversion apparatus according to the present invention. It is a block diagram which shows the structure of 2nd Embodiment of the image data converter which concerns on this invention. In order to explain the operation of the second embodiment of the image data conversion apparatus according to the present invention, FIG. It is a schematic diagram which shows the conventional image data conversion method which produces | generates reverse conversion image data from original progressive image data. It is a schematic diagram which shows the conventional image data conversion method which produces | generates progressive image data from original progressive image data.

Explanation of symbols

DESCRIPTION OF SYMBOLS 1 Sequence detection circuit 2 Switch 3 Normal progressive conversion circuit 4, 5 Telecine signal progressive conversion circuit 6, 46, 57 Selection circuit 41-43, 51-53 Frame memory 44, 54 Memory control part 45 Interpolation circuit 55 Motion detection circuit 56 Motion Compensation circuit

Claims (3)

Image data that converts interlaced image data of 60 fields / second obtained by converting film image data of 24 frames / second converted from a film source into an electrical signal by 2-3 pull-down method into progressive image data of 60 frames / second In the conversion device,
The interlaced image data for at least three frames of the film image data is written, and the odd field data and the even field data in the interlaced image data are synthesized for each frame of the film image data to have the same image quality as the film image data. A frame memory controlled to output the inverse transformed image data of
In the interlaced image data, two consecutive fields in which the first frame of the film image data is converted and three consecutive fields in which the second frame of the film image data is converted are combined. Sequence detection means for detecting which of the five fields present in the one sequence the interlaced image data is sequentially input with five fields as one sequence;
A control signal for sequentially writing the interlaced image data for at least three consecutive frames of the film image data in the frame memory and selectively reading the written interlaced image data based on the detection result of the sequence detecting means Memory control means for sequentially reading out the inversely transformed image data from the frame memory by:
Based on the detection result of the sequence detection means, the two inversely transformed image data read out continuously from the frame memory are combined and interpolated at a rate at which each image data component changes by 2/5. Interpolation means for generating frame image data;
Based on the detection result of the sequence detection means, one of the inversely transformed image data read from the frame memory is included in a position corresponding to any one field in the one sequence, and is generated by the interpolation means The four interpolated frame image data obtained are included in positions corresponding to the other four fields in the one sequence, respectively, and the time series of the inversely transformed image data and the interpolated frame image data included in the one sequence A selection means for selectively taking out from the frame memory and the interpolation means and outputting as the progressive data so that the components of the inversely transformed image data change by 2/5 each;
An image data conversion device comprising: Image data that converts interlaced image data of 60 fields / second obtained by converting film image data of 24 frames / second converted from a film source into an electrical signal by 2-3 pull-down method into progressive image data of 60 frames / second In the conversion device,
The interlaced image data for at least three frames of the film image data is written, and the odd field data and the even field data in the interlaced image data are synthesized for each frame of the film image data to have the same image quality as the film image data. A frame memory controlled to output the inverse transformed image data of
In the interlaced image data, two consecutive fields in which the first frame of the film image data is converted and three consecutive fields in which the second frame of the film image data is converted are combined. Sequence detection means for detecting which of the five fields present in the one sequence the interlaced image data is sequentially input with five fields as one sequence;
A control signal for sequentially writing the interlaced image data for at least three consecutive frames of the film image data in the frame memory and selectively reading the written interlaced image data based on the detection result of the sequence detecting means Memory control means for sequentially reading out the inversely transformed image data from the frame memory by:
A motion for detecting a motion vector between the two inversely transformed image data for each of the two inversely transformed image data continuously read from the frame memory based on the detection result of the sequence detecting means. Detection means;
Based on the detection result of the sequence detection means, the mutual motion vectors detected by the motion detection means are internally divided at a rate at which each component changes by 2/5, and generated by this internal division. Motion compensation means for motion-compensating and interpolating the inversely-transformed image data with two motion vector components as motion vectors for each of them, and averaging the obtained two motion compensation interpolation data to generate interpolated frame image data;
Based on the detection result of the sequence detection means, the one inversely transformed image data read from the frame memory is included in a position corresponding to any one of the one sequence, and is generated by the motion compensation means. The four interpolated frame image data are included in positions corresponding to the other four of the one sequence, respectively, and the time-series order of the inversely transformed image data and the interpolated frame image data included in the one sequence Selecting means for selectively taking out from the frame memory and the motion compensation means and outputting as the progressive data so that the components of the inversely transformed image data sequentially change by 2/5,
An image data conversion device comprising: The motion detection means divides the inversely transformed image data into a plurality of blocks, detects a motion vector for each of the blocks,
The motion compensation means generates two motion compensation block data from the two motion vectors generated by dividing the motion vector at a rate at which each component changes by 2/5, and the motion compensation block data Are interpolated block data, and the interpolated block data synthesized for the entire frame is the interpolated frame image data.
The image data conversion apparatus according to claim 2, wherein

Images